Considering Justin Verlander on Short Rest

Justin Verlander has never started a game on short rest in the way we conventionally think about short rest starts. Back in 2011, Verlander started Game 1 of the ALDS against the Yankees, but the game was suspended after one inning; he then started Game 3 of that series on two days’ rest. In 2017, Verlander started Game 1 of the ALDS against Boston, pitching six innings and throwing 99 pitches. He followed that start up with 2.2 innings of relief pitching after three days’ rest. Neither appearance really fits what we might expect to see from Verlander today when he starts on three days’ rest after pitching seven shutout inningsĀ on Friday.

Even if we add a history of Verlander making starts on short rest, given the very small number of starts and the huge time gap between them, it wouldn’t be especially relevant. Similarly, that Dallas Keuchel just fared poorly on short rest against the Cardinals probably also doesn’t have much bearing on what Verlander manages. The same is true for other pitchers over the last dozen years. There just isn’t a great body of evidence to tell us what we might see over the course of one game. What we do know is that Verlander is pitching a day earlier than he normally would, and that probably has some effect on his body and arm.

For the sake of argument, let’s say that pitching on short rest makes Verlander slightly weaker as a pitcher. We don’t know this to be the case, but it is certainly possible. And we can make another assumption that Verlander being slightly weaker means that his fastball won’t be as hard as it would be typically. Verlander’s fastball averaged 94.6 mph this season and in his start on Friday, it averaged 94.5 mph. If we wanted to know how Verlander might pitch on short rest, we could look at starts where his average fastball velocity was something less than his season-long average.

There were 10 Verlander starts this season where his fastball averaged 94.4 mph or lower. Here’s how he fared in those outings compared to the rest of his starts in which his fastball averaged at least 94.5 mph or more:

Justin Verlander with Less Velocity
High Velo Start Low Velo Start Difference
Starts 24 10 NA
IP/G 6.2 6 -0.2
AVG FA Velocity 94.8 93.7 -1.1
K% 38.3% 28.0% -10.3
BB% 4.1% 7.1% 3.0
HR/9 1.44 1.48 -0.04
FIP 2.96 4.09 -1.13
ERA 2.49 2.82 -0.33

In 10 starts with diminished velocity, Verlander’s strikeout rate went down significantly, his walks shot up, and his FIP was merely above average instead of great, with a slight downturn in ERA. He also got two fewer outs on average. Four of the low-velocity starts were in April, but even if we remove those starts we see similar results:

Justin Verlander with Less Velocity Since April
High Velo Start Low Velo Start Total Low Velo Starts Since April
Starts 24 10 6
IP/G 6.2 6.0 6.0
AVG FA Velocity 94.8 93.7 94.1
K% 38.3% 28.0% 28.6%
BB% 4.1% 7.1% 6.4%
HR/9 1.44 1.48 1.51
FIP 2.96 4.09 4.00
ERA 2.49 2.82 3.03

In any given game, a range of outcomes are fairly reasonable for Verlander, but if his velocity is off a tick, we should expect a little bit worse performance. When the options are Jose Urquidy plus the bullpen and then Verlander in Game 5 on regular rest if needed versus a short-rest Verlander and a full rest Gerrit Cole in Game 5 if needed, the potential drop off for Verlander in Game 4 is probably worth it given Cole’s presence. The move could backfire today and still be the right choice, but it isn’t fair to expect normal-rest Verlander in tonight’s game.

We hoped you liked reading Considering Justin Verlander on Short Rest by Craig Edwards!

Please support FanGraphs by becoming a member. We publish thousands of articles a year, host multiple podcasts, and have an ever growing database of baseball stats.

FanGraphs does not have a paywall. With your membership, we can continue to offer the content you've come to rely on and add to our unique baseball coverage.

Support FanGraphs

Craig Edwards can be found on twitter @craigjedwards.

newest oldest most voted

Interesting. Good research here, Craig.